Roese and Summerville’s meta-analysis of nine articles (11 data sets) found that romance is the third-most-cited regret domain, accounting for 14.8% of regrets behind education (32.2%) and career (22.3%), and that across regrets generally, those of action are more intense in the short term while those of inaction persist longer and intensify over time. A separate Northwestern/Kellogg study of 370 adults ages 19 to 103 found that 44% of women reported some romance-related regret (versus 19% of men), with those regrets prominently including someone they had feelings for but never pursued — the canonical “one that got away” — and never entering a particular relationship. There is no direct survey asking people who confessed romantic feelings and were rejected whether they regret the disclosure, so the action-regret rate of roughly 25% is a derived proxy, not a measured figure: it is anchored to the documented temporal pattern in which action regrets fade while inaction regrets endure.
Daniel Pink’s World Regret Survey, which has gathered more than 26,000 regrets from people in 134 countries, found that inaction regrets outnumber action regrets roughly two to one overall — a gap that widens to about three to one by people’s 50s — with connection and romance among the most consistently cited durable regret themes. This ratio is consistent with Gilovich and Medvec’s foundational temporal model, which holds that unexplored counterfactuals — “what if I had said something?” — remain open-ended and available to ruminate over, while action regrets are constrained by actual outcomes that eventually lose their emotional charge. Rejection, even when painful, forecloses the uncertainty; silence does not.
The main caveat is that the inaction-regret rate is anchored to the female rate from a single study conducted before app-mediated dating became dominant. Using the blended male-and-female rate from the same study (approximately 30 to 32%) would narrow the regret delta to around 0.05 to 0.07, still inaction-dominant but less dramatically so. The action-regret rate of 25% is estimated from indirect evidence rather than directly measured by a survey asking people who confessed romantic feelings whether they regret having done so. Both primary academic sources are more than 15 years old, and the emotional costs and norms around disclosure may differ for younger cohorts who navigate rejection through digital mediation. The directional finding that romantic inaction produces more durable regret than romantic action is among the most replicated results in regret research.